Cohort Profile
Cohort Profile: The DynaHEALTH consortium –
a European consortium for a life-course
bio-psychosocial model of healthy ageing
of glucose homeostasis
Sylvain Sebert,
1,2,3†Estelle Lowry
,
1,2†Nicole Aumu¨ller,
4Mercedes G Bermu´dez,
5Lise G Bjerregaard,
6Susanne R de Rooij,
7Maneka De Silva,
8Hanan El Marroun,
9Nadine Hummel,
10Teija Juola,
1Giacomo Mason,
11Daniela Much,
10Elena Oliveros,
12Stavros Poupakis,
11Nina Rautio,
1,2Phillipp Schwarzfischer,
4Evangelia Tzala,
8Olaf Uhl,
4Cornelieke van de Beek,
13Florianne Vehmeijer,
14Juan Verdejo-Roma´n,
15,16Niko Wasenius,
17,18Claire Webster,
19Leena Ala-Mursula,
1Karl-Heinz Herzig,
20,21Sirkka Keina¨nen-Kiukaanniemi,
1Jouko Miettunen,
1,22Jennifer L Baker,
6,23Cristina Campoy,
5Gabriella Conti,
11Johan G Eriksson,
17,18,24Sandra Hummel,
10Vincent Jaddoe,
25,26Berthold Koletzko,
4Alex Lewin,
27Maria Rodriguez-Palermo,
28Tessa Roseboom,
29Ricardo Rueda,
12Jayne Evans,
18Janine F Felix,
25Inga Prokopenko,
3Thorkild IA Sørensen
21,30and
Marjo-Riitta Ja¨rvelin
1,2,8,31*
1Centre for Life Course Health Research, Faculty of Medicine, University of Oulu, Finland,
2Biocenter
Oulu, University of Oulu, Finland,
3Department of Genomics of Complex Diseases, School of Public
Health, Imperial College London, UK,
4Division of Metabolic and Nutritional Medicine, Dr. von Hauner
Children’s Hospital, Ludwig-Maximilians Universita¨t Mu¨nchen, Munich, Germany,
5Department of
Paediatrics, School of Medicine, University of Granada, Spain,
6Center for Clinical Research and
Prevention, Bispebjerg and Frederiksberg Hospital, The Capital Region, Copenhagen, Denmark,
7Department of Clinical Epidemiology, Biostatistics & Bio informatics, Amsterdam University Medical
Centre),
8Department of Epidemiology and Biostatistics, MRC-PHE Centre for Environment and Health,
School of Public Health, Imperial College London, UK,
9Generation R Study Group, Department of
Pediatrics, Department of Child & Adolescent Psychiatry, Erasmus University Medical Center,
Rotterdam, the Netherlands,
10Institute of Diabetes Research, Helmholtz Zentrum Mu¨nchen, and
Forschergruppe Diabetes, Klinikum rechts der Isar, Technische Universita¨t Mu¨nchen, Neuherberg,
Germany,
11University College London, UK,
12Abbott Nutrition, Spain,
13Department of Obstetrics &
Gynaecology, Amsterdam University Medical Centers, The Netherlands,
14Generation R Study Group,
Department of Epidemiology, Erasmus University Medical Center, Rotterdam, the Netherlands,
15Mind,
Brain and Behavior Research Centre (CIMCYC), University of Granada, Spain,
16Department of
Experimental Psychology, Psychological Processes and Speech Therapy, Universidad Complutense de Madrid,
VCThe Author(s) 2019. Published by Oxford University Press on behalf of the International Epidemiological Association. 1051
This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted reuse, distribution, and reproduction in any medium, provided the original work is properly cited.
International Journal of Epidemiology, 2019, 1051–1051k doi: 10.1093/ije/dyz056 Advance Access Publication Date: 10 April 2019 Cohort Profile
17
Department of General Practice and Primary Health Care, University of Helsinki, Helsinki, Finland,
18Folkha¨lsan Research Center, Helsinki, Finland,
19Beta Technology, UK,
20Research Unit of Biomedicine,
Department of Physiology & Biocenter of Oulu, University of Oulu, Oulu, Finland,
21Department of
Gastroenterology and Metabolism, Poznan University of Medical Sciences, Poznan, Poland,
22Medical
Research Center Oulu, Oulu University Hospital and University of Oulu, Finland,
23NovoNordisk
Foundation Centre for Basic Metabolic Research, Section of Metabolic Genetics, Faculty of Health and
Medical Sciences, University of Copenhagen, Denmark,
24National Institute for Health and Welfare,
Finland,
25Generation R Study Group, Department of Pediatrics, Department of Epidemiology, Erasmus
University Medical Center, Rotterdam, the Netherlands,
26Harvard Medical School, USA,
27Department
of Medical Statistics, London School of Hygiene and Tropical Medicine, UK,
28Laboratorios Ordesa,
Spain,
29Academic Medical Centre, The Netherlands,
30Department of Public Health, Section of
Epidemiology, Faculty of Health and Medical Sciences, University of Copenhagen, Denmark and
31Department of Life Sciences, College of Health and Life Sciences, Brunel University London, UK
*Corresponding author. Professor and Chair in Lifecourse Epidemiology, Director of Postgraduate Research Degrees, Scientific Director of Northern Finland Birth Cohort Study, Department of Epidemiology and Biostatistics, School of Public Health, Faculty of Medicine, St. Mary’s campus, Imperial College London, W2 1PG, UK. E-mail: m.jarvelin@imperial.ac.uk
†These authors contributed equally to the manuscript.
Editorial decision 1 March 2019; Accepted 2 April 2019
Why was the collaboration set up?
DynaHEALTH is a research consortium funded by the European Commission through the Horizon 2020 research programme. It was established to help solve the societal challenge of an ageing population and the associated bur-den of non-communicable diseases related to obesity and type 2 diabetes (T2D). The consortium brings together European researchers and datasets to build an empirical model of unhealthy ageing, with a longitudinal perspective, in which causal, bi-directional, mediating and confounding factors operate at a multi-dimensional level (Figure 1).
The unhealthy ageing pathways often link altered adi-posity in early life, early-onset obesity, T2D and the further accumulation of other chronic physical and mental non-communicable diseases in older ages. At each stage of the life-course there is potential to intervene either on sus-pected biological causes via classical clinical approaches, or other plausible causative pathways1via integrated pub-lic health interventions (Figure 1). Successful interventions in early life have the potential to reduce subsequent invest-ments in later life. Scientific collaborations in epidemiology and public health such as DynaHEALTH are built on a long-standing tradition of collecting prospective data. Moving on to the era of open sciences policy,2meta-data analysis, the FAIR data principle (Findable, Accessible, Interoperable and Reusable) and the General Data Protection Regulation (GDPR), we must direct this public legacy in such a way as to better inform both policy makers and practitioners on complex patient and public health issues. Essentially, we must operate data to move beyond
association studies and explore how the psychosocial fac-tors, usually classified as confounders in medicine (Figure 2), can be analysed to help their operationalization and integration into healthcare programmes. Despite no apparent consensus in the literature on a single definition of psychosocial health, DynaHEALTH is referring to the following WHO definition3as a guiding principle: ‘a state of wellbeing in which every individual realises their own potential, can cope with the normal stresses of everyday life, can work productively and fruitfully, and is able to make a contribution to their community’.4,5
Conceptual framework of DynaHEALTH: The rela-tionships between psychosocial factors, glycaemic health4,5 and healthy ageing, including a reduced risk of T2D and cardiovascular diseases, may be conceptualizedin several different ways. It is an integral part of the DynaHEALTH consortium to develop these concepts and translate them into corresponding analytical study designs compatible with available data (Figure 2).
At the core of DynaHEALTH is the well-established ob-servational and likely causal relationship between stages of disease development: firstly between adiposity and the risk of deteriorating glycaemic health and eventual T2D,6–8and secondly between T2D and the risk of non-communicable diseases including stroke, coronary heart diseases, dementia and Alzheimer’s disease. All of these (pre-)clinical stages of disease development may share genetic, biological, lifestyle and psychosocial causes, but for both stages considerable knowledge gaps remain concerning action and impact on policies. We lack knowledge on identification of specific
1051a International Journal of Epidemiology, 2019, Vol. 48, No. 4
causal factors, the pathway(s) they operate through and their life-course aspects (Figure 2).
While considering these relationships, a number of hy-potheses may be posed about the role of psychosocial fac-tors throughout the life-course from the foetal period to old age.
• Unidirectional causality hypothesis: Adverse
psychoso-cial factors play a causal role in the development and worsening of adiposity, or of particular types of adipos-ity, with different relationships to glycaemic health and risk of T2D and subsequent morbidities.
• Pleiotropy and interaction hypothesis: In the case of
cau-sality in such life-course pathways, it can be hypothesized that causal reactions to one or a set of psychosocial fac-tors reflect underlying commonality influencing the clini-cal outcome in the life-course process. This can be analysed by exploring how the psychosocial factors may modify or contribute to the core relations between adipos-ity, glycaemic health, risk of T2D and the later onset of cardiovascular diseases, so that each of these relationships becomes stronger when the individuals are exposed to ad-verse psychosocial factors.
Figure 1. Illustration of the conceptual framework for unhealthy ageing from early life to old age. From a life-course perspective, it suggests that clini-cal events accumulate or lead to each other, and co-occur with factors that either can or cannot be modified. (T2D¼type 2 diabetes, NCD¼non-com-municable disease). (Reproduced from https://www.dynahealth.eu/with permission).
Figure 2. Physiological, lifestyle and psychosocial factors and stressors acting on life-course pathways linking altered adiposity in early life to non-communicable diseases at older ages.
International Journal of Epidemiology, 2019, Vol. 48, No. 4 1051b
• Bi-directional causality hypothesis: The relationships
between adiposity, glycaemic health, T2D risk and cardiovascular diseases are worsening one single or set of psychosocial factor(s) in such a way that vicious cycles of deterioration are promoted.
• Critical period hypothesis: There are specific time
peri-ods within the life-course during which psychosocial fac-tors have a greater impact on these unhealthy ageing pathways.
• Biological conversion hypothesis: The effects of
psycho-social factors can be ‘transformed’ into causal biological effects. In this case, we want to identify the persisting structural effects during early life and the functional mechanisms, e.g. by epigenetic regulation of gene expres-sion or changes in the metabolite profiles.
• Gene–environment hypothesis: The genetic variation
be-tween individuals may modify the transformation of psy-chosocial factors into biological effects.
The impact on future public health recommendations and new technologies strongly depends on the capacity to test this set of hypotheses. This requires large statistical power or the development of a specific study design that is enabled by building sustainable and targeted datasets through consortia such as DynaHEALTH.
Who is in the consortium?
To date, there is no single longitudinal study with a suffi-cient density of data and long-term follow-up to allow a life-course study of unhealthy ageing via changes in adult glycaemic health. Nonetheless, there is a wealth of some-what scattered, prospective studies with complementary designs which can be leveraged to study the dynamic deter-minants of life-long glycaemic health.
The DynaHEALTH consortium is a repertoire of hu-man studies with longitudinal design where key variables have been inventoried for meta-analysis or triangulation of evidence to test specific epidemiological concepts. The data we are analysing are from two main types of study design: prospective longitudinal surveys and randomized controlled trials (RCT). The consortium currently consists of 20 studies (Table 1) with data from eight European Countries (Figure 3) on up to 1 368 699 participants (Figures 4). The oldest living participants are from the Helsinki Birth Cohort Study (HBCS) born in 1934 in the Finnish city of Helsinki, whereas the youngest were born in April 2018 in the NIGOHEALTH RCT in Granada, Spain. These 20 studies consist of 12 general population cohorts9–20 of which six have a focus on later life and ageing with data beyond 65 years of age.9–11,17–19
Table 1. List of cohorts and randomized controlled trials participating in the DynaHEALTH consortium and their contacts
Abbreviation Cohort full name Contact
HBCS Helsinki Birth Cohort Study johan.eriksson@helsinki.fi
CSHRR Copenhagen School Health Record Register Jennifer.Lyn.Baker@regionh.dk
DFBC Dutch Famine Birth Cohort t.j.roseboom@amc.uva.nl
DCD Danish conscription database elme@sund.ku.dk
CPC Copenhagen Perinatal Cohort elme@sund.ku.dk
CIHVR Copenhagen Infant Health Visitor Records Jennifer.Lyn.Baker@regionh.dk NFBC1966 Northern Finland Birth Cohort 1966 NFBCprojectcenter@oulu.fi NFBC1986 Northern Finland Birth Cohort 1986 NFBCprojectcenter@oulu.fi
RS Rotterdam Study m.a.ikram@erasmusmc.nl
OULU1935 Oulu cohort study NFBCprojectcenter@oulu.fi
OULU1945 Oulu cohort study NFBCprojectcenter@oulu.fi
POGO German GDM Postpartum Outcomes in Women with Gestational Diabetes and their Offspring
sandra.hummel@lrz.uni-muenchen.de
GENERATION R Generation R Study v.jaddoe@erasmusmc.nl or
j.felix@erasmusmc.nl
PREOBE Excellence Project PREOBE ccampoy@ugr.es
Abbreviation Randomized controlled trials full name Contact
CHOP Childhood Obesity Program Berthold.Koletzko@med.uni-muenchen.de
RADIEL Finnish Gestational Diabetes Prevention Study johan.eriksson@helsinki.fi COGNIS A Neurocognitive and Immunological Study of a
New Formula for Healthy Infants
Maria.Rodriguez@ordesa.es
WOMB WOMB kids t.j.roseboom@amc.uva.nl
NIGOHEALTH Nutritional Intervention during Gestation and Offspring health Ricardo.Rueda@abbott.com This table has been adapted from https://www.dynahealth.eu/with permission.
1051c International Journal of Epidemiology, 2019, Vol. 48, No. 4
Three high-risk cohorts21–23 follow populations of off-spring born to mothers with a risk of gestational diabetes (GDM), either via high pre-pregnancy body mass index (BMI), previous history of GDM or other known risk fac-tors. Five RCTs focus24–28on pre-conception, prenatal and early-life interventions to improve maternal health and child development.
How often have they been followed up?
Altogether there are 17 time periods in which data has been prospectively collected from the pre-conception pe-riod until age 80 years of age (seeFigure 4for study time-lines). In addition, most follow-up time periods have data from at least two cohorts to allow replication or the use ofimputation across studies. The cohort-specific descriptions of follow-ups are provided inSupplementary File 1, avail-able asSupplementary dataat IJE online.
What has been measured?
The DynaHEALTH Consortium offers the potential to ex-tend our knowledge about the childhood origin of adult metabolic diseases and focus on key exposures during preg-nancy and early life and glycaemic, cardiovascular and metabolic outcomes in later life.
The consortium data harmonization policy is following a research-based focus where only variables defined in template-based analytical plans are being proposed for har-monization. The set of common harmonized data for
Figure 3. Map of studies participating in DynaHEALTH (sample size in brackets) by country. The size of the circle indicates the relative size of the study. Red arrows and stars show the participating centres of the multinational CHOP study.
International Journal of Epidemiology, 2019, Vol. 48, No. 4 1051d
DynaHEALTH has followed these steps adapted from Rolland et al.29
• Identification of the research questions that the
harmo-nized data set is required to answer.
• Identification of the high-level data concepts required to
answer those questions as described in the conceptual framework of DynaHEALTH.
• Assessment of data availability for data concepts. • Inventory of sets of pre-harmonized data due to
collabo-ration in previous consortiums.
• Development of analytical plans.
• Development of harmonized data following the FAIR
data management principles.
Retrospective harmonization of data, especially the psy-chosocial factors, can be a costly and sometimes an impos-sible process. In all cases analytical procedures shall account for the source of heterogeneity in inference which is addressed by the consortium via suitable meta-analytical processes and external replication.
In many cases, the cohort studies have been able to link with national, country-specific registers. These include databases such as population, hospital, education and em-ployment registers, health visitor records, hospital and
school records. This gives us the opportunity to use objec-tive data alongside self-reported responses. It also enables us to obtain more information than collected from the sur-veys, such as clinical diagnosis, hospital visits, medication use and cause of death. We also use information from these registers to conduct attrition analyses in the case of partici-pants lost during follow-up. In summary,Tables 2 and3 provide an overview of data available within each cohort. The co-ordinating team in DynaHEALTH has created, and regularly updates,a detailed inventory showing available data within each cohort, time points and method of collec-tion. This allows researchers to easily identify other data sources they could use to strengthen their study results, test a hypothesis or methodology, take a life-course approach by using data at earlier or later time points or investigate historical or cultural trends. A detailed inventory of what has been measured in each cohort can also be accessed on the DynaHEALTH website (http://dynahealth.eu/).
Pregnancy
A wealth of data is available from early life, as 16 of the study populations were established during pregnancy or even during the pre-conception period, following a cohort
Figure 4. DynaHEALTH datasets by size (in brackets), type, length and timing across the life-course, from the oldest study (top) to the youngest (bot-tom), separately for epidemiological longitudinal data and randomized controlled trials. White circles represent follow-up points, stars represent im-plementation of intervention and ‘R’ represents register-based datasets. Abbreviations are given inTable 1. (Reproduced from https://www. dynahealth.eu/ with permission).
1051e International Journal of Epidemiology, 2019, Vol. 48, No. 4
T able 2. Data availab le in general population-based studies (冑 ¼ data available, x ¼ data not available, -¼ no data collection at this time point) Indicato rs of inter est HBCS CSH RR DFBC DCD CPC CIH VR NFBC1 966 NFBC1 986 RS OU LU1 935 OU LU1945 PO GO German GDM GenR PRE OBE Pr egnan cy Anthro pometr ic mea sures - - - Blood sampl es - - - Gestat ional diabetes - - - Socio -econo mic indicator s - - - Heal th beh aviours - - - Ch ildho od (bir th to 12 y) Anthro pometr ic mea sures - - Growth mod elling - - Blood sampl es - - Dev elopmen tal milesto nes - - Early nutritio n - - Fam ily lifestyle informa tion - - Ad olescen ce (13– 18 y) Anthro pometr ic mea suremen ts - - - - -Blood sampl es - - - - -Socio -econo mic indicator s - - - - -Heal th beh aviours - - - - -Ad ulthoo d (18y þ ) Anthro pometr ic mea suremen ts - - --Blood sampl es - - - --Soci o-eco nomi c indi cator s - - --Heal th beh aviours - - -
--International Journal of Epidemiology, 2019, Vol. 48, No. 4 1051f
design or randomized clinical intervention (Figure 4). Maternal anthropometric measures such as height and weight have been recorded at various time points through-out pregnancy and in some studies at delivery, allowing the calculation of gestational weight gain. Blood samples have been taken from the mother during pregnancy and from the umbilical cord at birth enabling measures such as glu-cose, insulin or cardiovascular markers to be obtained. Additionally, extraction of DNA has enabled genotyping and DNA methylation arrays. Nuclear magnetic resonance (NMR)- or mass spectrometry-based metabolomics are now also available in a number of cohorts. Questionnaires were administered in some cohorts during pregnancy to collect social and demographic data such as work-related and household information. Health behaviours included smoking and alcohol use during pregnancy.
Childhood
Anthropometric information such as weight and height is particularly dense throughout early childhood as this was collected as part of routine practice within the health and welfare clinics in many European countries. Calculation of the BMI has allowed modelling of growth curves and ena-bles subsequent extraction of growth traits such as age at adiposity peak and rebound, and peak weight and height velocity. In addition, the health visitor records are particu-larly beneficial in obtaining early-life exposures such as age of achievement of common motor and cognitive
developmental milestones and breastfeeding duration. Some cohorts have also collected biological samples during this time period and more recently established studies in-clude detailed body composition measurements such as body fat percentage and skinfold thickness. Many studies include questionnaire responses in relation to a host of life-style information such as dietary intake, physical condition and activity, sleep duration and quality, parental smoking and alcohol use, parental occupations and maternity leave.
Adolescence
Height, weight and BMI are readily available for partici-pants during adolescence across the majority of cohorts, al-though measurements are less regular than in childhood. Some cohorts have collected biological samples that have allowed the inclusion of epigenetic and metabolomics data in collaborative analyses. Social information has been col-lected by questionnaire, primarily related to the social sta-tus of the parents. However, questions have been asked of the participants about their own smoking, alcohol and drug use, how they spend their leisure time and their typi-cal diet. In addition, some cohorts are linked to registered data on school performance.
Adulthood and old age
Anthropometric measures are readily available at many stages of adulthood, as well as blood samples. These have
Table 3. Data available or being collected for clinical trials in DynaHEALTH (冑¼ data available, x¼ data not available, - ¼no data collection at this time point)
Indicator of interest CHOP RADIEL COGNIS WOMB NIGOHEALTH
Pregnancy Anthropometric measures Blood samples Gestational diabetes Socio-economic indicators Health behaviours Childhood (birth to 12 y) Anthropometric measures Growth modelling Blood samples Developmental milestones Early nutrition
Family lifestyle information
Adolescence (13–18 y)
Anthropometric measurements - - -
-Blood samples - - -
-Socio-economic indicators - - -
-Health behaviours - - -
-1051g International Journal of Epidemiology, 2019, Vol. 48, No. 4
been used to derive a host of indicators including common cardio-metabolic biomarkers such as glucose, insulin and lipid levels and have been used to derive epigenetic and metabolomics information. Almost all cohorts have a vast array of psychosocial variables from questionnaires and national registers. Extensive information is available on employment and work-related information such as type of occupation, hours of work, employment history and in-come. Educational level and occupational training of the participant and their parents is also reported in many cohorts. We can obtain a wealth of information relating to family life, such as marital status, number of children, housing situation and living environment and their changes over time. Many cohorts have collected a range of back-ground information relating to lifestyle and health behav-iours including diet, physical activity, smoking and alcohol consumption. For old-age populations especially, measures of metabolic traits and cognition have been collected.
What has DynaHEALTH found? Key findings
and publications
A full list of publications arising from the DynaHEALTH ac-tion can be found on the project website (https://www.dyna health.eu/publications-map). DynaHEALTH aims to operationalize a data-driven approach to provide evidence for the bio-psychosocial pathways of unhealthy ageing asso-ciated with alteration of glycaemic health. This should also be examined in the context of strong health inequalities and complex transgenerational issues operating from the pre-conceptional period onwards. In support of others, we have identified several metabolic alterations in mothers with obe-sity and GDM in comparison with normal weight mothers, associated with offspring health, including changes in DNA methylation,30gene expression31and/or later metabolic out-comes.32,33From an age-related perspective, DynaHEALTH research has also contributed evidence to support links be-tween glucose metabolism, psychosocial factors and the age-ing process.34,35
As described in the above sections it is now essential to use and model the data to explore the nature of observed associations as a pre-requisite of informed decisions for prevention, intervention and policy making.36 For exam-ple, we have found little evidence for causality between maternal BMI during pregnancy andthe child’s risk of obe-sity. Rather, it is explained by genetic transmission of BMI variants interacting with a stressful, obesogenic environ-ment.37This is supported by a further study showing that risk scores based on genetic variants linked to specific bio-logical pathways influence body fat development from early life onwards. This study found an association be-tween a genetic risk score based on adult BMI, and BMI at
adiposity peak during infancy and abdominal fat measures at the age of 6 years.38
Ongoing research by the consortium is supporting the joint effects of bio-psychosocial factors on glucose metabo-lism.4We have also established an opportunity to change the trajectory of an individual from childhood adiposity to T2D in later life.39
Strengths and weaknesses: how does
DynaHEALTH offer a unique opportunity
and what are the main challenges faced
by the consortium?
DynaHEALTH exemplifies the wealth of prospective data collected in Europe that can be harnessed to enhance our understanding of healthy ageing by modelling the relation-ship between glycaemic health and psychosocial factors throughout the life-course. When combined, these data of-fer immense potential to inform future health policy in Europe. The data are organized to enable direct replication under collaborative agreements within the consortium and a number of observations can also be meta-analysed. While sample size allowing statistical power is deemed essential for robust evidence-based strategies, it is also important to combine study designs to validate findings under different statistical assumptions. DynaHEALTH’s additional strength is to include data from RCTs. Finally, the DynaHEALTH consortium includes both longitudinal birth cohorts and ageing cohorts from the same geographical lo-cation. This is the case in Northern Finland (NFBC1966/86 and Oulu 1935/45), the Rotterdam area (Rotterdam Study and Generation R Study) and in Copenhagen (CSHRR).
The critical mass of data, expertise and long-term col-laboration brought together in DynaHEALTH offers a sig-nificant resource. However, the main challenge faced by the consortium is how best to combine the characteristics of the cohorts involved. The cohorts were established for their own individual purposes before being brought to-gether under this project and the methods of data collec-tion have thus not been standardized a priori across the consortium. Therefore, some consideration is required for the transferability of the statistical models and there are similar challenges in harmonizing the data. In addition, this is an international project and therefore there are dif-ferences between studies and countries in technology, ques-tionnaire data and bio-specimen collection methods, terminology and diagnosis definitions, country-specific measurement techniques and ethical requirements.
However, the consortium has made significant progress in overcoming these challenges and the overarching opportunity for DynaHEALTH is that all studies provide rich data on similar key exposures and the outcome measures of interest.
International Journal of Epidemiology, 2019, Vol. 48, No. 4 1051h
Currently, there is no single cohort with data available from pconception to old age including information re-lating to both biological and psychosocial measures in rela-tion to health outcomes. This consortium provides unique access to a number of studies ranging over different time periods encompassing different life stages, that will enable us to use a life-course approach to model the risk of poor glycaemic health and T2D, and to better understand the dynamics of how this will change in response to other fac-tors throughout the life stages.
The collaborating cohorts include participants from eight European countries representing general, high-risk, obese and diabetic populations. This broad range of populations enables evaluation of the consistency of results and thus pro-vides greater generalizability of consortium findings. We also have access to expert collaborators from academia and indus-try with expertise in life-course epidemiology, developmental biology, genetics and epigenetics, metabolomics, biostatistics, clinical nutrition, health care, brain imaging, econometrics and European policy and knowledge management.
The key value of DynaHEALTH is that we foresee the well-established, strong collaboration being used as a plat-form for many further efforts and continuing beyond the Horizon 2020 programme. Thus, this consortium may pro-vide valuable assistance to investigators planning new cohort studies in terms of study design and the addition of new ideas, providing advice about data collection and manage-ment. DynaHEALTH will also offer an example of assumption-based modelling of the dynamic relations of the psychosocial and metabolic factors in the pathways of adi-posity -> glycaemic health –> T2D risk –> risk of cardiovas-cular disease –> unhealthy ageing throughout the life course.
Can I get hold of the data? Where can I find
out more?
The project is co-ordinated by the Centre for Life Course Health Research and the Northern Finland Cohort Project Centre at the University of Oulu in Finland. Further details on DynaHEALTH are available from the website: www. dynahealth.eu. The studies are approved by the local insti-tutional review boards. Written informed consent has been obtained for participants. There is no central repository for the data and each participating cohort has its own policy for data sharing.
The DynaHEALTH project legacy will be a collabora-tive action that will invite researchers with longitudinal life-course data to engage with us via the website (https:// www.dynahealth.eu/contact). To ensure continuity, the relevant section of the website and a light governance will remain to bring sustainability to the consortium and sup-port the testing ofits scientific concept.
Supplementary data
Supplementary dataare available at IJE online.
Funding
This project received funding from the European Union’s Horizon 2020 research and innovation programme under grant agreement No 633595.
Acknowledgements
Cohort-specific acknowledgements are listed in Supplement 2, avail-able asSupplementary dataat IJE online.
Conflict of interest: None declared.
References
1. Vandenbroucke JP, Broadbent A, Pearce N. Causality and causal inference in epidemiology: the need for a pluralistic approach. Int J Epidemiol 2016;45:1776–86.
2. Open Science Policy Platform. https://ec.europa.eu/research/ openscience/index.cfm? pg¼open-science-policy-platform (7 May 2018, date last accessed).
3. WHO Mental Health 2014. http://www.who.int/features/fact files/mental_health/en/ (18 December 2017, date last accessed).
4. Lowry E, Rautio N, Karhunen V, Miettunen J, Ala-Mursula L, Auvinen J. Understanding the complexity of glycaemic health—
DynaHEALTH profile in a nutshell
• DynaHEALTH exemplifies the wealth of prospective data collected in Europe that can be harnessed to enhance our understanding of healthy ageing by modelling the relationship between glycaemic health and psychosocial factors throughout the life-course.
• DynaHEALTH includes data on approximately 1.3 million subjects within 20 cohorts in eight European countries. Collectively, data spans the life-course from pre-conception through childhood and adult-hood into older age.
• Each individual study will adhere to its own protocol but generally we have repeated data collections dur-ing pregnancy and very early childhood. A number of studies have continued follow-up visits through to middle or old age. Other studies beginning in middle age have repeated measures relating to ageing.
• Data have been collected via clinical examinations including biological samples, brain scan images and questionnaire data relating to psychosocial varia-bles. Many studies also include linkage to national registers.
• Further details are available from the DynaHEALTH website: www.dynahealth.eu
1051i International Journal of Epidemiology, 2019, Vol. 48, No. 4
systematic bio-psychosocial modelling of fasting glucose in middle-age adults; a DynaHEALTH study. Int J Obes (Lond) 2018; doi: 10.1038/s41366-018-0175-1. [Epub ahead of print].
5. Danaei G, Finucane M, Lu Y, Singh G, Cowan M, Paciorek C. National, regional and global trends in fasting plasma glucose and diabetes prevalence since 1980: systematic analysis of health examination surveys and epidemiological studies with 370 country-years and 27 million participants. Lancet 2011;378: 31–40.
6. Horikoshi M, Beaumont R, Day FR et al. Genome-wide associa-tions for birth weight and correlaassocia-tions with adult disease. Nature 2016;538:248–52.
7. Freathy RM, Mook-Kanamori DO, Sovio U et al. Variants in ADCY5 and near CCNL1 are associated with fetal growth and birth weight. Nat Genet 2010;42:430–35.
8. Sovio U, Kaakinen M, Tzoulaki I et al. How do changes in body mass index in infancy and childhood associate with cardiometa-bolic profile in adulthood? Findings from the Northern Finland Birth Cohort 1966 Study. Int J Obes 2014;38:53–59.
9. Osmond C, Kajantie E, Forse´n T, Eriksson J, Barker D. Infant growth and stroke in adult life: the Helsinki birth cohort study. Stroke 2007;38:264–70.
10. Baker J, Olsen L, Andersen I, Pearson S, Hansen B, Sørensen T. Cohort profile: The Copenhagen school health records register. Int J Epid 2009;23:656–62.
11. de Rooij SR, Roseboom TJ. The developmental origins of ageing: study protocol for the Dutch famine birth cohort study on age-ing. BMJ Open 2013;3:e003167.
12. Christensen GT, Molbo D, A¨ ngquist LH et al. Cohort profile: The Danish Conscription Database(DCD): a cohort of 728, 160 men born from 1939 through 1959. Int J Epidemiol 2015;44:432–40. 13. Bjerregaard LG, Rasmussen KM, Michaelsen KF et al. Effects of
body size and change in body size from infancy through child-hood on body mass index in adultchild-hood. Int J Obes (Lond) 2014; 38:1305–11.
14. Andersen LG, Holst C, Michaelsen KF, Baker JL, Sørensen TIA. Weight and weight gain during early infancy predict childhood obesity: a case-cohort study. Int J Obes 2012;36:1306–11. 15. Rantakallio P. The longitudinal study of the northern Finland
birth cohort of 1966. Paediatr Perinat Epidemiol 1988;2:59–88. 16. Ja¨rvelin M-R, Hartikainen-Sorri A-L, Rantakallio P. Labour
in-duction policy in hospitals of different levels of specialisation. Br J Obstet Gynaecol 1993;100:310–15.
17. Ikram MA, Brusselle GGO, Murad SD, van Duijn CM, Franco OH, Goedegebure A. The Rotterdam study: 2018 update on objectives, design and main results. Eur J Epidemiol 2017;32: 807–50.
18. Rajala U, Kein˚Nen-Kiukaanniemi S, Kivel˚ S-L. Non-insulin-dependent diabetes mellitus and depression in a middle-aged Finnish population. Soc Psychiatry Psychiatr Epidemiol 1997; 32:363–67.
19. Hirsso P, Rajala U, Hiltunen L et al. Association of low-insulin sensitivity measured by quantitative insulin sensitivity check in-dex with hair loss in 55-year-old men. A Finnish population-based study. Diabetes Obes Metab 2006;8:466–68.
20. Kooijman MN, Kruithof CJ, van Duijn CM et al. The Generation R Study: design and cohort update 2017. Eur J Epidemiol 2016;31:1243–64.
21. Hummel S, Much D, Rossbauer M, Ziegler AG, Beyerlein A. Postpartum outcomes in women with gestational diabetes and their offspring: POGO study design and first-year results. Rev Diabet Stud 2013;10:49–57.
22. Boerschmann H, Pflu¨ger M, Henneberger L, Ziegler AG, Hummel S. Prevalence and predictors of overweight and insulin resistance in offspring of mothers with gestational diabetes melli-tus. Diabetes Care 2010;33: 1845–49.
23. Berglund S, Garcı´a-Valde´s L, Torres-Espinola FJ et al. Maternal, fetal and perinatal alterations associated with obesity, over-weight and gestational diabetes: an observational cohort study (PREOBE). BMC Public Health 2016;16:207.
24. Koletzko B, von Kries R, Closa R et al. Lower protein in infant formula is associated with lower weight up to age 2 y: a random-ized clinical trial. Am J Clin Nutr 2009;89:1836–45.
25. Ro¨no¨ K, Stach-Lempinen B, Klemetti MM et al. Prevention of gestational diabetes through lifestyle intervention: study design and methods of a Finnish randomized controlled multicenter trial (RADIEL). BMC Pregnancy Childbirth 2014;14:70. 26. Laboratorios Ordesa. A Neurocognitive and Immunological
Study of a New Formula for Healthy Infants (COGNIS). https:// clinicaltrials.gov/ct2/show/NCT02094547 (7 May 2018, date last accessed).
27. van de Beek C, Hoek A, Painter RC et al. Women, their Offspring and iMproving lifestyle for Better cardiovascular health of both (WOMB project): a protocol of the follow-up of a multicentre randomized controlled trial. BMJ Open 2018;8:e016579. 28. Abbott Nutrition. Nutritional Intervention During Gestation and
Offspring Health. https://clinicaltrials.gov/ct2/show/NCT02285764.
29. Rolland B, Reid S, Stelling D et al. Toward rigorous data harmo-nization in cancer epidemiology research: one approach. Am J Epidemiol 2015;182:1033–38.
30. Sharp GC, Salas LA, Monnereau C et al. Maternal BMI at the start of pregnancy and offspring epigenome-wide DNA methylation: findings from the pregnancy and childhood epigenetics (PACE) consortium. Hum Mol Genet 2017;26: 4067–85.
31. Martino J, Sebert S, Segura MT et al. Maternal body weight and gestational diabetes differentially influence placental and preg-nancy outcomes. J Clin Endocrinol Metab 2016;101:59–68. doi: 10.1210/jc.2015-2590.
32. Santos Ferreira DL, Williams DM, Kangas AJ et al. Association of pre-pregnancy body mass index with offspring metabolic pro-file: Analyses of 3 European prospective birth cohorts. PLoS Med 2017;14:e1002376.
33. Wu¨rtz P, Wang Q, Niironen M et al. Metabolic signatures of birthweight in 18 288 adolescents and adults. Int J Epidemiol 2016;45:1539–50.
34. Rautio N, Varanka-Ruuska T, Vaaramo E et al. Accumulated exposure to unemployment is related to impaired glucose metabolism in middle-aged men: A follow-up of the Northern Finland Birth Cohort 1966. Prim Care Diabetes 2017;11:365–72. 35. von Bonsdorff MB, von Bonsdorff ME, Haanpa¨a¨ M et al.
Work-loss years among people diagnosed with diabetes: a reappraisal from a life course perspective. Acta Diabetol 2018;55:485–91. doi: 10.1007/s00592-018-1119-x.
36. Swanson SA, Tiemeier H, Ikram MA, Herna´n MA. Nature as a Trialist?: deconstructing the analogy between mendelian
International Journal of Epidemiology, 2019, Vol. 48, No. 4 1051j
randomization and randomized trials. Epidemiology 2017;28: 653–59.
37. Richmond R, Timpson N, Felix J et al. Using genetic variation to ex-plore the causal effect of maternal pregnancy adiposity: a Mendelian randomisation study. PLOS Med 2017;2414:e1002221. 38. Monnereau C, Vogelezang S, Kruithof CJ, Jaddoe V, Felix J. Associations of genetic risk scores based on adult adiposity
pathways with childhood growth and adiposity measures. BMC Genetics 2016;17:120.
39. Bjerregaard LG, Jensen BW, A¨ ngquist L, Osler M, Sørensen TIA, Baker JL. Change in overweight from childhood to early adult-hood and risk of type 2 diabetes. N Engl J Med 2018;378: 1302–12.
1051k International Journal of Epidemiology, 2019, Vol. 48, No. 4
